Genome-wide association studies (GWAS) have identified >100 common variants associated with body mass index (BMI) and obesity risk. Pathway and tissue expression analyses of nearby genes have provided strong evidence for a role of the central nervous system (CNS) in body weight regulation. However, these GWAS variants have small effects, are common, non-coding, intronic or intergenic, do not alter protein function and are typically not the true disease-causing variants. Yet, knowing the causal gene/variant is critical for the translation of a GWAS locus into new insights in the biology of body weight regulation. Thus, there is a clear need for more effective gene-discovery strategies to identify causal genes/variants, which in turn will facilitate the translation of gene discoveries into new biology. Therefore, we propose to screen the exome for rare (MAF<1%) coding single nucleotide variants (SNVs) associated with BMI and obesity risk using data from 1 million people, with the aim to expedite the pinpointing of causal SNVs (Aim 1). We will subsequently prioritize the identified coding SNVs based on their implications on protein function and enrichment in extremely obese cases (Aim 2). The top-ranked coding SNVs will be functionally characterized in human induced pluripotent stem cell (hiPSC)-derived cellular models (Aim 3). Specifically, in Aim 1, we will apply and customize methods optimized for mega-scale cohorts to perform the first 1 million-scale exome-wide association study leveraging exome-chip genotype data from two international collaborations; the GIANT consortium (N>525,000) and the UKBiobank (N~500,000). We have >90% statistical power to identify coding SNVs with a MAF as low as 0.02% that have clinically relevant effect sizes (>6 kg/allele (>13.2 lbs) for a 1.7m (5ft 7in) tall person). In Aim 2, we develop an analytical pipeline to prioritize the identified coding SNVs from Aim 1. The pipeline will annotate SNVs, quantify their intolerance with regard to impact on gene function, and identify the tissues that are affected the most. We will examine whether SNVs are enriched in extremely obese cases using unique study designs, including a discordant family study and a longitudinal study of longtime extreme obesity. A set of 10-15 prioritized coding SNVs will be functionally characterized in Aim 3. We will knock the respective coding SNV into hiPSC using CRISPR/Cas9 and differentiate these into the relevant cell type (e.g. neurons, adipocytes, beta cells, gut cells, hepatocytes). We will then investigate the impact of SNV on cellular and molecular obesity-relevant phenotypes to elucidate underlying biology. We are uniquely positioned to identify and functionally characterize rare coding SNVs for obesity. Such SNVs have the promise to disproportionally increase our understanding of the biology of obesity and may lead to new and more precise strategies for prevention and treatment of obesity, a field that has seen little innovation in the past 30 years.